Re: My First Tesla Coil

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: Jim Lux <jimlux-at-earthlink-dot-net> 

At 04:21 PM 5/13/2004 -0600, you wrote:
>Original poster: "Chris Fanjoy" <zappyman-at-hotmail-dot-com>
>   I've wanted to build one of these for several years now, but just never 
> got around to it. But finally, I'm now ready to build my first coil!

Go for it.. Don't expect your first coil to be perfect or 
wonderful.  There's a lot of "art" in building a coil, from figuring out 
how to wind secondaries, to improvising toroids for the top, to tuning, 
etc.  A heck of lot of fun.

>   I received the main ingredient in the mail today - a 10 lb spool of #27 
> AWG magnet wire. Upon seeing the immense size of the spool, I now realize 
> that I have more wire than I'll probably ever need! Not a problem though, 
> since an associate of mine also wishes to build a coil, and I'll probably 
> just share some with him. As for the other materials - high voltage 
> transformers, capacitors, large-diameter copper wire, etc - I have a 
> large stock of these items that I've collected, from the years I've spent 
> repairing consumer electronics.



>   I've got quite a mouthful of questions, but I'll try to summarize them 
> as briefly as possible:
>
>Consideration #1 : Number of coil turns / turns ratio
>
>   Most info I've collected from the web, suggests winding about 2-4 
> thousand turns of 20-24 AWG magnet wire, for the secondary of the coil. 
> The primary should be about 20-25 turns of heavier (<14 AWG) copper wire. 
> Is it theoretically possible, to use a turns radio of 1:100 (say, 
> primary=10 turns and secondary =1000 turns), to multiply the input 
> voltage by 100x? (assuming 100% efficiency, obviously it will be much 
> lower in real life). Secondly, how does one determine the proper number 
> of turns? And finally, can the primary and secondary windings be isolated 
> from each other by putting a larger-diameter piece of PVC pipe between 
> them? (to prevent arcing)

First... download one of the spreadsheets or java tools to do the 
calculations!  It will save you tons of time, because you can fiddle with 
the different parameters.

Turns ratio means nothing in the tesla coil context.  It's ratios between 
inductances and capacitances that determines stuff.

The latest guidance is to go for around 1000 turns or so, and make your 
secondary 4-5 times as high as it is in diameter.

The primary should be a flat spiral, typically in the 10-15 turns 
area.  Use bare (solid) wire or copper tubing and space the turns far 
enough apart that you can get some sort of clip on it to adjust the 
tuning.  Make the inside diameter of the spiral an inch or so bigger than 
the secondary diameter.

Make a provision so that you can change the height of the secondary above 
the primary (this lets you change the coupling between primary and secondary)




>Consideration #2 : High Voltage Transformer
>
>   Since I have literally dozens of old MOT's in my parts pile, it goes 
> without saying that this will be what I use. My first question is, are 
> the older (bigger) ones better? Size might be a consideration depending 
> on the enclosure I built this stuff into. And also, is it possible to use 
> two transformers (primaries wired in parallel, secondaries in series) in 
> order to get twice as much voltage to feed to the spark gap/capacitor? 
> Since two MOT's would draw more than 15 amps from the wall, some kind of 
> current limiting would be needed (custom-wound inductor?)

Yes, you can series to MOTs to get around 6-8kV.


>Consideration #3 : Tank Capacitor
>
>   It seems that everyone uses a tank cap comprised of dozens, or 
> hundreds, of small HV caps to get the required value and voltage. I have 
> a fair number of mylar caps (mostly from 70's and 80's color TV's) which 
> may do the job, but I'm wondering if microwave oven capacitors would 
> work. I have dozens of em', and most have values of 0.7 - 1.0 uF at up to 
> 2400VAC - each! Only thing is, they have bleeder resistors (typically 
> 10meg-ohms each) and might not withstand high operationg frequencies. And 
> what about high-voltage ceramic capacitors, are these any good?


Mylar has poor HF loss characteristics and is not suitable. MO capacitors 
are too big and are really designed to work at 60 Hz.

Your real options are these:
1) Beer bottles filled with salt water in a bucket filled with salt 
water.  Heavy, cheap, works fairly well.
2) the mass of capacitors (called a MMC), using polypropylene dielectric 
and construction designed for high peak currents.  Not all polypropylene 
caps are suitable.
3) Ceramic doorknob transmitting caps (or even, the non-transmitting kind, 
but they're more lossy and can't really take the current).  These aren't 
the greatest thing performance wise, but sometimes you can get a really, 
really good price (like a buck apiece for 2700 pF 40kV units) and then you 
might as well use them.  If you're getting up into spending $40-50+ on 
doorknob capacitors to get the required capacitance, it's time to consider 
the MMC route.  MUCH better performance and about the same cost.
4) Purpose built high rep rate HV pulse discharge capacitors (i.e. the 
Maxwell labs units). Wretchedly expensive brand new (hundreds of $), 
sometimes available surplus cheaply (although they have limited life.. so 
you want new surplus, not some old ratty used thing that was pulled at the 
end of its rated life)

#1 and #2 are your best options, unless you happen to score a deal with #3 
or #4.


>   Maybe that's enough questions for now. It will take some time to 
> assemble all the components, and work out the design. I look forward to 
> any advice you may have.

You'll also need a spark gap...

I'd start with three parallel lengths of copper pipe (3/4 or 1" pipe, 2-3 
inches long, F-F sweat on couplings would work nicely). Two get bolted down 
and connected to the two HV leads, space them about half the diameter of 
the pipe. The other is arranged like stacking logs, and can be moved to 
adjust the gap spacing.